Thought suppression failures in combat PTSD: A cognitive load hypothesisq
Deane E. Aikinsa,b,*, Douglas C. Johnsona,b, Jessica L. Borellic, David H. Klemanskic, Paul M. Morrisseyd,
Todd L. Benhamd, Steven M. Southwicka,b, David F. Tolina,e
aDepartment of Psychiatry, Yale University, 300 George Street, New Haven, CT 06511, USA
bNational Center for Posttraumatic Stress Disorder, Clinical Neuroscience Division, 950 Campbell Avenue/151E, West Haven, CT 06516, USA
cDepartment of Psychology, Yale University, 2 Hillhouse Avenue, New Haven, CT 06520, USA
d10th Mountain Division (LI), 10000 10th Mtn Div Dr, Fort Drum, NY 13602, USA
eInstitute of Living, 200 Retreat Avenue, Hartford, CT 06106, USA
a r t i c l e i n f o
Received 24 July 2008
Received in revised form
10 June 2009
Accepted 11 June 2009
a b s t r a c t
The present study investigated the relation between thought suppression of emotionally neutral content
[i.e., Wegner’s (1994) ‘‘white bear’’], incidental traumatic thought intrusion, and skin conductance
responses in combat-related Posttraumatic Stress Disorder (PTSD). Participants included service
members who either: a) had PTSD following an Operation Iraqi Freedom deployment; b) were free of
psychiatric diagnosis following deployment (Combat Equivalent), or c) were pre-deployed and without
psychiatric diagnosis (Pre-Deployed). PTSD Service Members reported the greatest intrusion of combat
thoughts during the suppression task and demonstrated a post-suppression rebound effect with
a neutral thought. Non-specific skin conductance responses indicated that the suppression task was
related to similar levels of increased sympathetic activity for both the PTSD and Pre-Deployed groups,
whereas the Combat Equivalent group showed no increased activation during thought suppression.
Intrusive traumatic thoughts combined with failures in neutral thought suppression may be a conse-
quence of increased cognitive load in PTSD.
? 2009 Elsevier Ltd. All rights reserved.
Intrusive thoughts are an important aspect of Posttraumatic
Stress Disorder (PTSD). Horowitz (1986) suggested that traumatic
events must be assimilated intothe schema of a person’s life events.
This assimilation process may be inhibited, due to the intense
unpleasant arousal of the traumatic event. Assimilation failures are
posited to cause ‘‘leaks’’ of traumatic material into consciousness in
the form of intrusive thoughts. In accord with recent clinical
models of PTSD (Resick & Schnicke, 1992) corrective integration of
a traumatic event into the life schema is thought to decrease
intrusive thoughts and alleviate PTSD. However, despite its theo-
retical significance, the cognitive mechanism of thought intrusion
has yet to be fully understood.
Information-processing models of thought suppression provide
insight into the mechanisms that may underlie intrusive thoughts
in PTSD. Wegner’s (1994) prototypical ‘‘white bear’’ paradigm
requires individuals to observe the frequency of a novel target
thought (i.e., the ‘‘white bear’’) during task periods thateither allow
for (‘‘monitor’’) or discourage (‘‘suppress’’) its emphasis (Wegner,
Schneider, Carter, & White,1987). Most individuals are successful in
actively suppressing a target thought, yet some experience an
increase in target thought frequency once the suppression task has
ended (a ‘‘rebound’’ effect during a second monitoring task; Wen-
zlaff & Wegner, 2000). According to the ironic process theory,
thought suppression is supported bytwo cognitive mechanisms: an
effortful search system that seeks to engage in thought content
other than the target thought and a non-effortful monitoring
system that flags instances when the search system has failed
(Wegner,1992). The non-effortful monitoring system operates with
automaticity and may continue when the effortful search system
has ended. In this manner, the monitoring system identifies the
increase in target thoughts following the suppression task, creating
the rebound effect.
Preliminary studies of thought suppression in PTSD indicate
rebound-related effects with trauma target thoughts. Increased
trauma-related thought rebound has been associated with PTSD
following sexual assault (Shipherd & Beck, 1999) and motor vehicle
qThis study was supported by the National Center for Posttraumatic Stress
Disorder, Clinical Neuroscience Division.
* Corresponding author at: National Center for Posttraumatic Stress Disorder,
Clinical Neuroscience Division, 950 Campbell Ave/151E, West Haven, CT 06516, USA.
Tel.: þ12039325711x5557; fax: þ12039373481.
E-mail address: firstname.lastname@example.org (D.E. Aikins).
Contents lists available at ScienceDirect
Behaviour Research and Therapy
journal homepage: www.elsevier.com/locate/brat
0005-7967/$ – see front matter ? 2009 Elsevier Ltd. All rights reserved.
Behaviour Research and Therapy 47 (2009) 744–751
or distress about the intrusive thought was reported in PTSD partic-
ipants. Similar rebound effects have been reported with PTSD and
treatment-seeking traumatized non-PTSD individuals (Beck, Gud-
mundsdottir, Palyo, Miller, & Grant, 2006) and traumatized college
student samples (e.g., Davies & Clark, 1998; Rosenthal, Cheavens,
Lynch, & Follette, 2006; Tull, Gratz, Salters, & Roemer, 2004).
Traumatic thought content may directly lead to suppression
rebound effects. Alternatively, evidence of rebound effects with
non-trauma thoughts may be reflective of PTSD-related deficits in
thought suppression ability. In an effort to determine content
specificity in PTSD-related rebound effects, MVA survivors were
asked to monitor and suppress both trauma and non-trauma
neutral personal thoughts (Shipherd & Beck, 2005). Results indi-
cated onlya rebound effect with the trauma thought. This finding is
consistent with a model of emotional processing which posits that
recent exposure to trauma content may prime PTSD individuals to
inhibit responses to incongruent emotional content (Litz, Orsillo,
Kaloupek, & Weathers, 2000). In this regard, thinking about the
trauma in close proximity to a non-trauma suppression task may
decrease access to non-trauma thoughts that results in an absence
of a rebound effect. Thus, engaging PTSD participants without
a discussion of traumatic thought contentprior to a neutral thought
suppression task would preclude priming to a trauma context.
Although a recent investigation of trauma and neutral thought
content attempted to examine this issue (Amstadter & Vernon,
2006), its use of a different suppression methodology (i.e. an initial
baseline phase that did not introduce the target thought, which is
the usual practice with an initial monitoring phase) confounds
suppression with cuing effects manifest in the initial monitoring
task. As a result, these findings are rendered in comparable to those
reviewed above (see Wenzlaff & Wegner, 2000, for review of
methodologies). Thus, the scope of thought suppression rebound
effects in PTSD populations remains to be clarified.
An additional aspect of thought suppression that has yet to be
considered in the PTSD literature is the role of cognitive load on
thought suppression failures. Wegner (1992) proposed that
suppression failures occur when task demands (increased task
complexity) overwhelm the effortful search system. In studies that
utilized a variety of manipulations (time constraints, dual tasks),
suppression failures were observed in samples of healthy partici-
pants (Wegner & Erber, 1992). Thus, at times of high demand,
effortful thought suppression –for a neutral target thought- fails
due to competition for cognitive resources (Wenzlaff & Wegner,
2000). We hypothesize that PTSD individuals may engage in an on-
going effortful traumatic thought suppression process in a manner
similar to that described by Horowitz (1986). This traumatic
suppression process would limit the available cognitive resources
for PTSD individuals to engage in other concurrent tasks. Thus, it
might be expected that PTSD individuals are prone to trauma
thought suppression failures during periods of relatively modest
cognitive demand, such as a neutral thought suppression task.
Autonomic sympathetic activity should also be concordant with
the cognitive effort required to suppress thoughts. For example,
Lacey and colleagues (Lacey, Kagan, Lacey, & Moss, 1963) found
increased Skin Conductance Level during a variety of tasks. In
a similar manner, non-specific Skin Conductance Responses (NS-
SCR), spontaneous skin conductance responses observed over some
discrete period of time, have been found to increase in frequency
during completion of complex tasks (Munro, Dawson, Schell, &
Sakai, 1987). Increases in NS-SCR frequency has been conceptual-
ized as an indicator of heightened autonomic activation that is
associated with an effortful allocation of attentional resources
(Jennings, 1986). Yet, psycho-physiological monitoring has only
been included in one study of PTSD thought suppression (Beck
et al., 2006), which used heart rate, skin conductance level, and
frontalis EMG as measures of autonomic arousal in response to
traumatic thoughts. Only minimal group differences in these
measures were found, such that the PTSD group had elevated EMG
during the suppression task relative to the control group. The
authors conceptualized their use of the physiological measures as
indices of hyper arousal associated with processing traumatic
thought content. In this manner, they speculated that the lack of
physiological differences between PTSD and non-PTSD groups was
possibly related to the shared degree of distress over their trauma
and treatment-seeking status, regardless of PTSD diagnosis. Inter-
estingly, there were no reported task-related physiological effects,
despite the fact that both groups demonstrated a thought rebound
effect. To date, the use of NS-SCR as a measure of autonomic arousal
associated with task effort has yet to be utilized in a thought
suppressionparadigm. Thus, tothe extent that thought suppression
is an effortful strategy, concomitant increases in NS-SCR should be
found during the suppression task.
To our knowledge, the present paper is the first to investigate
thought suppression effects in individuals with combat-related
PTSD. The specificity of a suppression-related rebound effect was
tested with a standard neutral thought task (‘‘the white bear’’).
Evidence of a rebound effect with a standard neutral thought would
further support the conceptualization of a dysregulated thought
suppression mechanism in PTSD. Mood and emotional responses to
the thought tasks were recorded in an effort to relate these factors
as potential covariates to suppression ability. Second, we wanted to
explore the consequence of neutral thought suppression to the on-
going frequency of trauma thought intrusions. As suggested above,
the added effort to suppress a novel, affectively neutral thought is
predicted to deplete the already tapped cognitive resources in PTSD
individuals and result in an increase of incidental intrusive trauma
thoughts during the suppression task. Finally, we expected to see
an increase in NS-SCR activity during the suppression task relative
to the monitoring conditions, as an indicator of increased auto-
nomic activation associated with increased cognitive effort. To the
extent that PTSD Service Members are engaged in a dual suppres-
sion task (both trauma and white bear thoughts) and thus requiring
more cognitive resources than non-PTSD groups, they should show
greatest NS-SCR activity.
The participants were 43 right-handed active-duty male Service
Members, between the ages of 19 and 37, based at Fort Drum, New
York. All were members of the 10th Mountain Light Infantry Divi-
sion (LI). Members of this LI Division were among the first troops
deployed in Operation Iraqi Freedom (see Table 1 for demographics
and diagnostic data). For those participants who served in Iraq, the
average length of time back from Iraq following their first 12-
month deployment was 8 months (SD¼5.3).
Participants were comprised of three groups: 14 who were
combat-exposed during deployment to Iraq and subsequently
developed PTSD (PTSD group); 14 who were combat-exposed
during the Iraq deployment and did not meet criteria for current
diagnosis of PTSD or any other Axis I diagnosis (Combat Equivalent
group); and 15 who were yet to be deployed to Iraq, not combat-
exposed, and did not have a current Axis I diagnosis (Pre-Deployed
group). All participants were free from medical complications and
were not taking any prescription or over-the-counter medications.
All PTSD participants met criteria for current PTSD via the
Clinician Administered PTSD Scale (CAPS; Blake et al., 1995). The
control participants did not meet criteria for current Axis I or Axis II
disorders as determined by the Structured Clinical Interview for
D.E. Aikins et al. / Behaviour Research and Therapy 47 (2009) 744–751 745
DSM-IV Axis I (SCID-P; First, Spitzer, Gibbon, & Williams,1995) and
the Structured Clinical Interview for DSM-IV Axis II Personality
Disorders (SCID-II; First et al., 1995). Four trained clinical psychol-
ogists administered the SCID and CAPS assessments. These
psychologists had extensive experience in assessing anxiety
disorders (including PTSD) and completed an assessment training
program at the National Center for PTSD Clinical Neurosciences
Division prior to the start of this study. In cases where diagnosti-
cians disagreed, the diagnoses of record were based upon team
Participants were recruited from Fort Drum via media adver-
tisements, informational flyers, and announcements made during
Battalion meetings looking for combat-deployed or pre-deployed
research volunteers. During a pre-screening phone interview,
participants were excluded if they had any medical conditions
that would affect the neuroimaging procedures included as an
aspect of the larger investigation, were taking any over-the-
counter or prescription medications,
significant psychotic symptoms as defined by DSM-IV, had current
substance abuse or dependence, or had any organic mental
disorders. Out of 71 phone interviews, we excluded 14 (20%)
individuals for MR-compatibility reasons (handedness, claustro-
phobia, medical metal implants, shrapnel, or dental implants), and
9 (13%) for current (including psychiatric) medication use. The 43
participants were selected from the remaining pool of 50 potential
participants based on availability. Furthermore, all participants
were forewarned that they would be excluded from the study if
they were found to be actively using alcohol or illicit drugs as
determined by urine drug screen or breathalyzer test adminis-
tered during the data collection. No participant failed the drug and
alcohol screens during the study days.
the Institute of Living at Hartford Hospital. Each individual partici-
pated in a larger study involving 2 weekend days of testing, divided
into ten 90-minute sessions that were randomized in a Latin-square
time of the day. Other sessions included functional and structural
neuroimaging, and diagnostic and assessment interviews. The
current study is reporting on a thought suppression session and the
results of the psychological diagnostic interview sessions.
The following clinician administered diagnostic scales and self-
report metrics were administered to each participant.
Clinician Administered PTSD Scale (CAPS)
The CAPS is a structured clinical interview measuring the
frequency and intensity of the 17 DSM-IV symptoms of PTSD and
the 8 associated symptoms. The CAPS also contains five global
rating questions regarding the impact of symptoms on social and
occupational functioning, improvement since previous assessment,
and overall validity and severity of reported symptoms. The CAPS
has an excellent test-retest reliability (.90–.98) and internal
consistency (.94). Totalseverityscore on theCAPS has been strongly
correlated with the Mississippi Scale for PTSD (.91) and the PK scale
of the MMPI-2 (.77). In the validated sample, CAPS total severity
scores ?65 had adequate sensitivity (.84), specificity (.95), and
efficiency (.89) against PTSD diagnosis based on the Structured
Clinical Interview for DSM-IV (SCID) (Blake et al., 1995).
Structured Clinical Interview for DSM-IV (SCID-P)
The SCID (First et al., 1995) was used to assess comorbid diag-
nostic status for all participants. The SCID is a widely used instru-
ment with acceptable psychometric properties.
Combat exposure scale
Participants completed the Combat Exposure Scale (CES-D;
Keane et al., 1989), a 7-item questionnaire based on Vietnam-
of exposureto combat-related experiences involving danger, injury,
or death. Published reports indicate that the measure has internal
cross-cultural validity (Askari, 2005; Keane et al., 1989).
Beck Depression Inventory (BDI)
The BDI (Beck, Ward, Mendelson, Mock, & Erbaugh, 1961) is
a 21-item self-report measure of severity of depression that has
high reliability and validity (Beck, Steer, & Carbin, 1988).
At the beginning of the thought suppression session, partici-
pants were briefly trained tocomplete a semantic differential mood
rating scale (see Osgood, Suci, & Tannenbaum, 1957) in which 15
pairs of contrasting mood adjectives are located at either end of
a vertical line and participants are asked to draw a mark corre-
sponding to where their mood state falls in relation to the two
words. Items are constructed to reflect two principle mood factors,
valence (i.e., happydsad, gladdangry, positivednegative) and
arousal (i.e, peacefuldagitated, relaxeddanxious, calmdtense).
Responses to items are totaled for the valence and arousal factors.
Interpretation of the factor scores is that a value of 0 reflects
neutrality. Negative scores reflect high arousal or negative valence.
Positive scores reflect low arousal or positive valence.
Thought Description Questionnaire (TDQ)
The TDQ (Tolin, Abramowitz, Przeworski, & Foa, 2002) consists
of 5 items that ask about occurrences of the target thought during
the most recent 5-min period: (1) frequency of the thoughts
(response was a number); (2) time spent having the thought
(response was a percentage from 0 to 100); (3) pleasantness of the
thought (rated on a seven-point scale from 1, extremely pleasant, to
7, extremely unpleasant); (4) vividness of the thought (rated on
a seven-point scale from 1, not at all vivid, to 7, extremely vivid);
and (5) efforts tosuppressthe thought (rated on a seven-point scale
from 1, no effort, to 7, extremely strong effort).
Thought monitoring was recorded via a tape recorder placed on
a table in front of the participant. Each participant recorded the
Demographic summary for combat PTSD, combat equivalent, and Pre-Deployed
Years of Education
Years in the Service
Age, education, and years in service, multivariate F (6, 76)¼0.47, n.s.
D.E. Aikins et al. / Behaviour Research and Therapy 47 (2009) 744–751746
number of thoughts about a white bear with an event key marker
(Tally I Hand Model Tally Counter, ACCO Brands, Booneville, MS).
Skin Conductance Response was recorded with two Ag-AgCl elec-
trodes filled with an isotonic gel with 0.5% saline in a neutral base
(GEL101, BIOPAC Systems, Inc.) were attached via adhesive collars
to the medial phalanx of the palmar surface of the first two fingers
of the non-dominant hand. Signals were processed with BIOPAC
MP150 system attached to a PC laptop and sampled at 1 kHz. Non-
specific Skin Conductance Responses (NS-SCR) were defined as
events that were 0.05 microsiemens or larger (Andreassi, 2000)
and constrained within a 3 second latency window of both task
onset and a prior occurrence of a NS-SCR. All responses were
Participants were seated in front of a table with the tape
recorder and psycho-physiological recording apparatus, in a small,
sound-attenuated room. The experiment consisted of a 5-min
baseline of non-activity, a 3-min thought monitoring practice
period, and three 5-min trials: monitor 1, suppression, and monitor 2
(rebound). In order to minimize reactivity, the experimenter left the
room during each trial. During each trial, frequency of thoughts
about a ‘white bear’ was collected via event-marking to indicate the
occurrence of the target thought and an audio tape recorder, which
recorded the verbal thought monitoring task.
After applying the skin conductance electrodes, the experi-
menter informed the participant to sit quietly with his eyes open
for five minutes, after which the task would begin. Following the
baseline, participants completed the semantic differential.
Each participant received the following verbal instructions:
‘‘During this study you will be asked to monitor your thoughts out
loud. Whatever thoughts you have are fine. Don’t censor yourself
or worry about what thoughts you have or don’t have. Just notice
them and let them pass. With every different thought you have,
press the event key. First, you will have a brief trial so that you can
practice monitoring your thoughts.’’ Participants practiced moni-
toring their thoughts for 3 min. The experimenter verified
participants’ compliance via event key markers; additional prac-
tice trials were used when necessary for the participant to
understand the task.
Trial 1 (monitor 1)
Next, participants were given instructions in the same manner
as the practice session: ‘‘Just like before, you will monitor out loud
everything that comes to mind. During this time, you may think
about anything that you like. For example, you may think about
a white bear, but you don’t have to. What you think about is up to
you. However, if you do happen to think about a white bear, please
press the event key.’’ The participants were left alone for 5 min.
After Trial 1, the participants reported their degree of suppression
effort, the amount of time spent thinking about the target thought,
and the valence and vividness of the thought on the TDQ. The
experimenter recorded the number on the event key marker on the
TDQ and reset the value to zero.
Trial 2 (suppression)
The participants were then given the following instructions:
‘‘In the next five minutes, please monitor your thoughts as you did
before, with one exception. This time, try NOT to think of a white
bear. Every time you think ‘white bear’ or have a white bear come
to mind, though, please press the event key.’’ The participants
were left alone for 5 min. After Trial 2, the participants again
reported their mental activity on the TDQ. The experimenter
recorded the number on the event key marker on the TDQ and
reset the value to zero.
Trial 3 (monitor 2)
The participants were then given the following instructions:
‘‘Just like before, you will monitor out loud everything that comes
to mind. During this time, you may think about anything that you
like. For example, you may think about a white bear, but you don’t
have to. What you think about is up to you. However, if you do
happen tothink about awhitebear, please press the event key.’’ The
participants were left alone for 5 min. After Trial 3, the participants
completed the same questionnaires. The experimenter recorded
the number on the event key marker.
Following Trial 3, the experimenter gave the participant a final
version of the TDQ that asked the participant to rate intrusive
thoughts about combat in Iraq during each trial of the experiment.
Following data collection, the audio tapes of the thought
monitoring tasks were transcribed. In a manner similar to other
PTSD thought suppression studies, all thoughts were coded as
either OIF combat-related, white bear-related, or ‘‘other’’ (see
Shipherd & Beck, 2005). Thought listing data were coded by two
trained independent raters. Raters were uninformed of the diag-
nostic status of the participants and met training criteria by
reaching 80% agreement on five consecutive practice cases (Clark,
Ball, & Pape, 1991). A total of 92% agreement was reached on
thought listing data with the remaining discrepancies (8%) resolved
by a third trained rater. Individual variability in the number of
thoughts listed across participants was controlled by calculating
the percentage of OIF- and white bear-related thoughts for each
participant for each phase.
A multivariate analysis of variance (MANOVA) revealed no
significant group differences in age, education, and years of military
service (see Table 1). There were no significant differences between
the groups on age and education, with the mean age across the
groups ranging from 24.14 to 25.47 years of age and the mean years
of education ranging from 12.43 to 12.80. The groups were also
similar with regard to ethnicity (see Table 1), with a little more than
half of the participants in all three groups composed of Caucasian
men, consistent with the demographics of the Army. Additionally,
chi-square analyses indicated there were no group differences in
PTSD symptom severity
Table 2 highlights group self-report profiles. All combat-
exposed participants met A1 and A2 criterion for an event during
OIF deployment. By design, a MANOVA indicated significant
symptom differences reported among the PTSD, Combat Equiva-
lent, and Pre-Deployed groups [F (10, 70)¼41.56, p<0.0001,
p¼0.86]. Univariate ANOVAs indicated group differences in
combat exposure, depressive symptoms, and total symptoms of
traumatic stress [CES, F (2, 40)¼167.37, p<0.0001, p¼0.89; BDI, F
(2, 40)¼12.99, p<0.0001, p¼0.39; CAPS total score, F (2,
D.E. Aikins et al. / Behaviour Research and Therapy 47 (2009) 744–751747
40)¼79.69, p<0.0001, p¼0.80].1Post-hoc analyses (with Bon-
ferroni adjustment p<0.05) indicated increased BDI scores in the
PTSD group relative to both the Combat Equivalent and Pre-
Deployed groups.2Similar post-hoc analyses indicated differences
in combat exposure only between Pre-Deployed and the two
deployed groups. Univariate ANOVAs for each respective cluster
resulted in significant group differences for re-experiencing (Cluster
B), F (2, 40)¼54.78, p<0.0001, p¼0.73, avoidance and numbing
(Cluster C), F (2, 40)¼59.54, p<0.0001, p¼0.75, and hyper arousal
(Cluster D), F (2, 40)¼57.31, p<0.0001, p¼0.74. Post-hoc analyses
of eachclusterindicated that the PTSD group hadsignificantly more
severe CAPS symptoms than the Combat Equivalent and Pre-
Deployed groups. More symptoms were reported in the Combat
Equivalent than the Pre-Deployed group for Cluster B and D.
An additional analysis was conducted on the CAPS data for
Criteria F (Distress, social impairment, and occupational impair-
ment) between PTSD and Combat Equivalent groups. A 2
(Group) ?3 (Criteria) MANOVA indicated a main effect for Group, F
(3, 24)¼19.85, p<0.0001, p¼0.71. The PTSD group had higher
scores than the Combat Equivalent group on measures of distress [F
(1, 28)¼63.70, p<0.0001, p¼0.71], social impairment [F (1,
28)¼34.67, p<0.0001, p¼0.57], and occupational impairment [F
(1, 28)¼14.36, p<0.001, p¼0.36].
Only two of the Service Members in the PTSD group met criteria
for a comorbid Axis I diagnosis (Dysthymic Disorder) that existed
prior to deployment, and none of the Service Members met criteria
for current substance abuse or dependence. Thirty-five percent of
the sample (n¼15) was randomly selected for an independent
rater to listen to the audio taped interviews and establish inter-
rater reliability of diagnoses. The kappa coefficient for PTSD diag-
nosis was .93.
White bear thoughts3
In a test of the first hypothesis, the percentage of white bear
thoughts was entered into a 3 (Group)?3 (Thought Condition)
mixed factor ANOVA, with thought condition as the repeated
measure. As seen in Fig. 1, a significant main effect for Thought
Condition, F (2, 39)¼4.63, p<0.02, p¼0.19, was modified by
a Group?Condition interaction, F (2, 40)¼3.76, p<0.03, p¼0.16.
For all participants, there was a significant reduction in the
frequency of white bear thoughts between the Monitor 1 and
Frequency of white bear thoughts remained unchanged from the
Suppression to the Monitor 2 condition for the Pre-Deployed and
Combat Equivalent Groups, F (1, 27)¼6.83, p<0.02, p¼.20, but
was at a frequency that was still lower than the Monitor 1 condi-
tion, F (1, 27)¼16.95, p<0.0001, p¼0.39. In contrast, PTSD
participants reported a significant increase in the percentage of
white bear thoughts from the Suppression to Monitor 2 condition, F
(1, 13)¼7.21, p<0.01, p¼.55. The percentage of white bear
thoughts during the Monitor 2 condition was significantly greater
for the PTSD group (M¼51.8%, SE¼3.0), than for the Combat
Equivalent (M¼18.1%, SE ¼3.0) or the Pre-Deployed groups
(M¼20.3%, SE¼3.0%), F (2, 40)¼4.98, p<0.012, p¼0.20.
Baseline self-reported arousal and valence ratings did not differ
across the three groups, F (4, 80)¼2.16, n.s. On the TDQ, PTSD
participants rated the white bear thoughts as more vivid than did
the Combat Equivalent and Pre-Deployed participants, F (2,
40)¼3.08, p<0.05, p¼0.13, (Bonferroni post-hoc adjustments
significant at alpha<0.05). There were no group differences in
reported pleasantness or suppression effort. Baseline arousal and
valence ratings and TDQ scores were also assessed as potential
covariates with thought intrusion across groups. These measures
had no significant effects on the thought intrusion results.
Iraq combat thoughts
In a test of our second hypothesis, percentage of intrusive
combat trauma thoughts found in PTSD and Combat Equivalent
groups were analyzed utilizing a 2 (Deployed Group)?3 (Thought
Condition) mixed factor ANOVA, with thought condition as the
Clinical summary for combat PTSD, combat equivalent, and Pre-Deployed
CAPS (Clinician Administered PTSD Scale); CES (Combat Exposure Scale); BDI (Beck
Depression Inventory); Different superscripts indicate significant differences at
Fig. 1. Evidence of a rebound effect for a standard neutral thought in PTSD service
members. Percentage of neutral thoughts significantly increases only in the PTSD
group during the monitor 2 condition. Error bars indicate SEM.
1All effect sizes were calculated using partial eta squared, the ratio of variance
accounted for by the independent variable over the sum of the variance accounted
for by the independent variable and the variance attributed to error.
2BDI scores for the two PTSD Service Members diagnosed with Dysthymic
Disorder were 28 and 24, respectively. The mean PTSD BDI score with these two
participants excluded was 8.09 (3.75), with a PTSD group maximum score of 13.
There were no changes in the tests of our hypotheses when these participants were
removed. The reported results include the two dysthymic PTSD participants.
3Analyses were similar with both the percentage frequency and event key
thought monitoring methods. The percentage data are reported to provide
comparisons to past PTSD studies. Further, analysis of absolute scores did not yield
D.E. Aikins et al. / Behaviour Research and Therapy 47 (2009) 744–751748
repeated measure. As seen in Fig. 2, a significant main effect for
Deployed Group, F (1, 26)¼5.19, p<0.03, p¼0.17, was modified by
a Deployed Group?Condition interaction, F (2, 25)¼4.10, p<0.02,
p¼0.14. When divided into the three task conditions, there was
only a significant difference in number of intrusive combat
thoughts in the Suppression condition between the PTSD and
Combat Equivalent groups, t (26)¼2.90, p<0.008. Additionally,
within-subjects contrasts indicated that although frequency of
combat thoughts remained the same between the two thought
monitor conditions for the PTSD participants, F (1,13)¼2.33, n.s.,
there was a significant increase in the frequency of intrusive
combat thoughts between the Monitorand Suppression conditions,
F (1, 13)¼4.87, p<0.05. No other comparisons were significant.
As per report on the TDQ, the PTSD Service Members rated Iraq
related combat thoughts as more vivid, t (26)¼3.24, p<0.003, and
reported more effort to suppress them, t (26)¼2.48, p<0.02, than
group differences in the pleasantness of the trauma thoughts,
t (26)¼1.46, n.s.
Therewere no deployed
Skin conductance response
An initial Univariate ANOVA indicated no differences in NS-SCR
frequency at baseline between the three groups, Univariate F (2,
40)¼0.43, n.s., p¼0.02. In a test of our third hypothesis, a signifi-
cantmultivariate main effect
(3, 38)¼81.29, p<0.0001, p¼0.87, was modified by a Group-
?Condition interaction, F (6, 78)¼3.17, p<0.008, p¼0.20 (See
Fig. 3). All groups showed a significant increase in NS-SCR
frequency from baseline to the thought tasks, all within contrasts
significant at p<0.0001. There were no significant differences in
NS-SCR frequency between PTSD and Pre-Deployed groups. In
considering the combination of PTSD and Pre-Deployed groups, the
number of NS-SCRs significantly increased from the Monitor 1 to
Suppression condition, F (1, 27)¼7.54, p<0.01, p¼0.22, and then
decreased from the Suppression to Monitor 2 conditions, F
(1,27)¼13.41, p<0.001, p¼0.33. There was no significant differ-
ence between NS-SCRs during the Monitor 1 and Monitor 2
conditions, F (1, 27)¼0.33, n.s., p¼0.01. In contrast, NS-SCRs in the
Combat Equivalent group remained unchanged throughout the 3
task conditions, F (2, 12)¼0.32, n.s., p¼0.05. The Combat Equiva-
lent group had fewer NS-SCRs during the Suppression condition
relative to both Pre-Deployed and PTSD groups, F (2, 40)¼7.33,
p<0.002, p¼0.27, all Bonferroni-corrected post-hoc comparisons
significant at alpha¼0.05.
Thought intrusion is a critical component of the PTSD diagnosis.
Yet, the information-processing mechanism of intrusive thoughts
remains to be fully explicated. In the PTSD literature, aberrations in
thought suppression may be specific to trauma content, or broadly
related to the suppression process itself. In the present study, we
found support for the latter: PTSD Service Members demonstrated
post-suppression rebound with a neutral ‘‘white bear’’ target
thought, whereas both non-PTSD Combat Equivalent and Pre-
Deployed Service Members did not. Further, we tested a cognitive
load hypothesis, suggesting that traumatic thought intrusions for
individuals with PTSD occur at times of relatively high cognitive
demand, such as a thought suppression task. We hypothesized that
PTSD Service Members engage in an on-going process of sup-
pressing unwanted combat-related thoughts, a partial reflection of
the lack of integration of this material into their personal schemas
(Horowitz, 1986). In this manner, directions to suppress a neutral
thought (the ‘‘white bear’’) would be most likelytocause a failure in
traumatic thought suppression. In support of this hypothesis, PTSD
Service Members reported an increase in traumatic thoughts
during the suppression task, whereas Combat Equivalent Service
Members did not. Finally, we hypothesized that the cognitive
demand of the neutral thought suppression task would be associ-
ated with increased ANS reactivity as measured by an increased
NS-SCR frequency. Results revealed that all Service Members
demonstrated an NS-SCR increase from resting baseline to the
thought tasks. However, PTSD and Pre-Deployed Service Members
demonstrated an additional NS-SCR increase during the suppres-
sion task, whereas, Combat Equivalent Service Members showed no
additional NS-SCR activity during thought suppression.
Our findings both support and extend the literature in several
ways. In a manner similar to studies of civilian samples using
trauma-specific thoughts, our PTSD Service Members demon-
strated a rebound effect with a standard neutral thought. In
contrast to past studies, this rebound effect was found in the
absence of any reported negative mood or valence changes during
the task. This evidence lends support to a conceptualization of
a more general PTSD-related information-processing anomaly that
may underlie difficulties with suppressing trauma-specific content.
The current findings are in contrast to past work that failed to
find evidence of post-suppression rebound effects with neutral
thoughts. Shipherd and Beck (2005) investigated both traumatic
and neutral thought suppression performance during the same
laboratory session. The flooding effects of trauma content on later
information-processing abilities has long been discussed in the
PTSD literature, particularly with regard to the trauma-specific
order effects for modified Stroop tasks (see McNally, Kaspi,
Fig. 2. Relative to non-PTSD combat equivalent exposed participants, PTSD individuals
report a significant increase in trauma thoughts during a neutral thought suppression
task. Error bars indicate SEM.
Baseline Monitor 1 Suppress Monitor 2
Fig. 3. Non-specific skin conductance responses increase during cognitive activity. An
additional increase in responses is indicated in PTSD and Pre-Deployed service
members. Error bars indicate SEM.
D.E. Aikins et al. / Behaviour Research and Therapy 47 (2009) 744–751749
Riemann, & Zeitlin,1990). In this regard, it is possible that both the
anticipation and facilitation of trauma-specific thoughts in that
study had an effect on subsequent tasks such that neutral thoughts
were not actively suppressed. In the present study, rebound effects
with a neutral thought were found when no mention of traumawas
made by the experimenter prior to or during the thought tasks, so
as to not facilitate focus on trauma. As such, future investigations
will need to make the distinction between individuals who
suppress a neutral thought with and without an additional explicit
trauma thought monitoring task.
An additional component of this study was examination of the
differential effects of cognitive demands on the three groups. We
hypothesized that cognitive load would contribute to performance
decrements in the PTSD participants. It has been shown that
manipulating task demands significantly affects thought suppres-
sion failures, such that increased task complexity gives rise to
intrusive thoughts (Wenzlaff & Wegner, 2000). In the present study,
we considered the neutral thought suppression task as relatively
high cognitive load, predicated on the notion that PTSD Service
Members engage in an on-going suppression of trauma-related
content. Wegner (1994) broadly conceptualized cognitive load as
multitasking, such that additional working memory tasks would be
hypothesized to disrupt suppression. As such, intrusive traumatic
thoughts may be evidenced during other demanding (yet non-
psychological assessments. In this manner, our finding supports
a conceptualization of PTSD in which general cognitive processing
ability is negatively affected by the on-going task demands of
processing traumatic thoughts.
This was the first study to use NS-SCR frequency as a measure of
autonomic activation that would correspond with cognitive load in
a thought suppression paradigm. Past research indicated NS-SCR
increases during conditions associated with greater cognitive
complexity (Munro, Dawson, Schell, & Sakai,1987). Consistent with
this work, all groups showed an NS-SCR frequency increase relative
to resting baseline. Further, both PTSD and non-PTSD Pre-Deployed
Service Members demonstrated an additional NS-SCR increase
during the suppression task. Yet, these NS-SCR increases were
found in the absence of similar increases in self-rated suppression
effort. We had expected PTSD participants would be more
cognitively active during the suppression task, as they were
hypothetically engaged in dual suppression tasks (i.e., spontaneous
suppression of trauma content and directed suppression of neutral
content). Instead, the Combat Equivalent group showed a unique
NS-SCR pattern that did not differentiate the three tasks (i.e.,
monitor, suppress, post-suppression monitor). It may be the case
that the NS-SCRs of the PTSD and Pre-Deployed groups are
reflective of a normative response pattern to the thought
The present study’s incorporation of combat equivalent Service
Members without psychiatric diagnoses allows for a consideration
of the individual differences that may contribute to resilience
(Masten, 2001; Southwick, Vythilingam, & Charney, 2005). In an
extrapolation of Horowtiz’s (1986) model with the present data, we
propose that the Combat Equivalent Service Members more
successfully integrated the traumatic combat experience into their
personal schemas than those with PTSD. As a result, the Combat
Equivalent group did not have to allocate cognitive resources
towards on-going traumatic thought suppression. For these
participants, the laboratory suppression task is not considered an
instance of high cognitive load, thus no significant increases in
trauma-related thoughts would be expected. In this manner, the
Combat Equivalent group was able to successfully suppress the
white bear thought without requiring additional sympathetic
engagement, as measured by NS-SCR. As these participants were
exposed tothe same combat trauma, were diagnosis-free, and most
importantly, reported no impairments from their trauma exposure,
it raises the possibility that they represent a select group who are
In the execution of this study, several possible limitations were
considered: 1) whereas the PTSDsymptom severityof our sample is
similar with that of the previous thought suppression PTSD data,
the duration of illness is far more acute. It may be possible that
additional thought suppression and mood effects are related to
chronic PTSD illness. 2) Our sample consisted entirely of men,
thereby preventing the investigation of potential relevant gender
effects. 3) Further, our study did not include a group that only
monitored through the three tasks. As such, we do not know if the
frequency of neutral thoughts would increase without the
suppression instructions. 4) Also, we utilized the Combat Exposure
Scale, which was based on the experiences of Vietnam-era
Veterans. Whereas our use of the CES allowed for a comparison of
the combat severity in the present work to the majority of other
combat-related PTSD research, recent measures, such as the
Deployment Risk and Resilience Inventory (King, King, & Vogt,
2003) more accurately reflect the combat events likely deployed in
recent desert warfare. Thus, our understandingof the combat stress
experienced by the deployed groups in the present study must take
this consideration into account. 5) Finally, we do not know if
increased mental effort in general or thought suppression per se is
related to increased intrusive thoughts.
In summary, findings from the current study have important
contributions to our understanding of PTSD as well as trauma
resilience. Combat-related PTSD was associated with a greater
incidence of combat-related traumatic thoughts during a neutral
thought suppression task, followed by a post-suppression rebound
effect with a neutral thought. This profile of findings has not been
found to date in the PTSD thought suppression literature, and
suggests a PTSD information-processing anomaly that is not
specific to trauma-related thoughts. Whereas a cognitive load
model may partially account for these results, we found no
evidence of PTSD-related increased sympathetic activity (measured
by NS-SCR frequency) during the thought suppression task. Rather,
the Combat Equivalent group successfully suppressed neutral
thoughts without any increase in traumatic thought intrusion or
NS-SCR increases. The unique performance from the Combat
Equivalent group provides preliminary evidence for a cognitive
process that may be indicative of resilient responses to traumatic
stress. Thus, thought suppression mechanisms may make impor-
tant contributions to a variety of outcomes following traumatic
The authors wish to thank the Service Members of the 10th
Mountain Division (LI) for their participation in this study.
Amstadter, A. B., & Vernon, L. L. (2006). Suppression of neutral and trauma targets:
implications for posttraumatic stress disorder. Journal of Traumatic Stress, 19,
Andreassi, J. L. (2000). Psychophysiology: Human behavior and physiological response
(4th ed.). Mahwah, NJ: Lawrence Erlbaum Associates Publishers.
Askari, S. (2005). Psychometric evaluation of combat exposure scale in Iranian war
veterans. Psychological Research, 7, 58–64.
Beck, A. T., Steer, R. A., & Carbin, M. G. (1988). Psychometric properties of the Beck
Depression Inventory: twenty-five years of evaluation. Clinical Psychology
Review, 8, 77–100.
Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory
for measuring depression. Archives of General Psychiatry, 4, 561–571.
D.E. Aikins et al. / Behaviour Research and Therapy 47 (2009) 744–751 750
Beck, J. G., Gudmundsdottir, B., Palyo, S. A., Miller, L. M., & Grant, D. M. (2006). Download full-text
Rebound effects following deliberate thought suppression: does PTSD make
a difference? Behaviour Therapy, 37, 170–180.
Blake, D. D., Weathers, F. W., Nagy, L. M., Kaloupek, D. G., Gusman, F. D.,
Charney, D. S., et al. (1995). The development of a clinician-administered scale,
PTSD scale. Journal of Traumatic Stress, 8, 75–90.
Clark, D. M., Ball, S., & Pape, D. (1991). An experimental investigation of thought
suppression. Behaviour Research and Therapy, 29, 253–257.
Davies, M.I., & Clark, D. M. (1998).Though suppressionproduces a rebound effectwith
analogue post-traumatic intrusions. Behaviour Research and Therapy, 36, 571–582.
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. (1995). Structured Clinical
Interview for the DSM-IV Axis I disorders-patient edition (SCID-I/P, Version 2.0).
New York: Biometrics Research Dept., New York State Psychiatric Institute.
Horowitz, M. J. (1986). Stress-response syndromes: a review of posttraumatic and
adjustment disorders. Hospital & Community Psychiatry, 37, 241–249.
Jennings, J. R. (1986). Bodily changes during attending. In M. G. H. Coles, E. Donchin,
& S. W. Porges (Eds.), Psychophysiology: Systems, processes, and applications (pp.
268–289). New York: Guilford Press.
Keane, T. M., Fairbank, J. A., Caddell, J. M., Zimering, R. T., Taylor, K. L., & Mora, C. A.
(1989). Clinical evaluation of a measure to assess combat exposure. Psycho-
logical Assessment, 1(1), 53–55.
King, D. W., King, L. A., & Vogt, D. S. (2003). Manual for the DeploymentRisk
and Resilience Inventory (DRRI): A collection of measures for studying
deployment – Related experiences of military veterans. Boston, MA: National
Center for PTSD.
Lacey, J. I., Kagan, J., Lacey, B. C., & Moss, H. A. (1963). The visceral level: situational
determinants and behavioral correlates of autonomic response patterns. In
P. H. Knapp (Ed.), Expressions of the emotions in man (pp. 161–196). New York:
International Universities Press.
Litz, B. T., Orsillo, S. M., Kaloupek, D., & Weathers, F. (2000). Emotional processing in
posttraumatic stress disorder. Journal of Abnormal Psychology, 109, 26–39.
McNally, R. J., Kaspi, S. P., Riemann, B. C., & Zeitlin, S. B. (1990). Selective processing
of threat cues in posttraumatic stress orders. Journal of Abnormal Psychology, 99,
Masten, A. S. (2001). Ordinary magic: resilience processes in development. Amer-
ican Psychologist, 56, 227–238.
Munro, L. L., Dawson, M. E., Schell, A. M., & Sakai, L. M. (1987). Electrodermal lability
and rapid performance decrement in a degraded stimulus continuous perfor-
mance task. Journal of Psychophysiology, 1, 249–257.
Osgood, C., Suci, G., & Tannenbaum, P. (1957). The measurement of meaning. Urbana,
IL: University of Illinois.
Resick, P. A., & Schnicke, M. K. (1992). Cognitive processing therapy for sexual
assault victims. Journal of Consulting and Clinical Psychology, 60, 748–756.
Rosenthal, M. Z., Cheavens, J. S., Lynch, T. R., & Follette, V. (2006). Thought
suppression mediates the relationship between negative mood and PTSD in
sexually assaulted women. Journal of Traumatic Stress, 19, 741–745.
Shipherd, J. C., & Beck, J. G. (1999). The effects of suppressing trauma-related
thoughts on women with rape-related posttraumatic stress disorder. Behaviour
Research & Therapy, 37, 99–112.
Shipherd, J. C., & Beck, J. G. (2005). The role of thought suppression in posttraumatic
stress disorder. Behavior Therapy, 36, 277–287.
Southwick, S. M., Vythilingam, M., & Charney, D. S. (2005). The psychobiology of
depression and resilience to stress: implications for prevention and treatment.
Annual Review of Clinical Psychology, 1, 255–291.
Tolin, D. F., Abramowitz, J. S., Przeworski, A., & Foa, E. B. (2002). Thought
suppression in obsessive-compulsive disorder. Behaviour Research and Therapy,
Tull, M. T., Gratz, K. L., Salters, K., & Roemer, L. (2004). The role of experiential
avoidance in posttraumatic stress symptoms and symptoms of depression,
anxiety, and somatization. Journal of Nervous and Mental Disease, 192,
Wegner, D. M. (1992). You can’t always think what you want: problems in the
suppression of unwanted thoughts. Advances in Experimental Social Psychology,
Wegner, D. M. (1994). White bears and other unwanted thoughts: Suppression,
obsession, and the psychology of the mental control. New York: Guilford Press.
Wegner, D. M., & Erber, R. (1992). The hyperaccessibility of suppressed thoughts.
Journal of Personality and Social Psychology, 63, 903–912.
Wegner, D. M., Schneider, D. J., Carter, S. R., & White, T. L. (1987). Paradoxical effects
of thought suppression. Journal of Personality and Social Psychology, 53, 5–13.
Wenzlaff, R. M., & Wegner, D. M. (2000). Thought suppression. Annual Review of
Psychology, 51, 59–91.
D.E. Aikins et al. / Behaviour Research and Therapy 47 (2009) 744–751 751